Filter Out False Positives
When monitoring any application for anomalies, the challenge can be that each individual parameter being monitored will stray outside of its expected normal range of operation, but unless other parameters are also showing stress this may not be important. But when a series of different parameters all start to show a change in the same period of time this could indicate a potential issue.
Simple systems that create alerts when any single parameter or series of parameters hit defined thresholds are only useful up to a point. If the flow of alerts of inconsequential pieces of information is too great, then they can mask really important data. If the volume of alerts is to great you can’t spot important alerts quickly enough, then the alerting process just doesn’t work. Most people refer to this white noise of inconsequential alerts as false positives.
Nastel’s AutoPilot products have technology expressly designed to avoid the false positive issue.Autopilot sends alerts when a business view detects that a fact it is monitoring has been changed, and it evaluates several other meta data internally before an alert is generated. For example, if a business view has sent an alert and has been restarted without any modification and the fact remains constant, the false alert suppression logic will prevent an alert from getting generated again.
Additionally, if an error has been detected on a lower level object, while the state of the higher- level object is unknown, alert will not be issued for the lower level object. For example, if a channel is retrying, but the corresponding Queue Manager and Node status is unknown, an alert will not be issued for the channel, but the appropriate sensors for Queue Manager or Node will send the corresponding alert.
The result is that Nastel uses innovate technology to ensure your monitoring team is only alert to true issues, and false positives are cleaned out of the alerting process. Nastel is the leading provider of monitoring and management for IBM MQ.